US 20080198456 A1
Chemically-bonded retarder stacks are provided in this disclosure. A first organic layer having a first molecular orientation is chemically welded to a second organic layer having a second molecular orientation. The first and second molecular orientations are crossed when the organic layers are laminated. The first and second organic layers may be polycarbonate films, which can be welded together using a suitable solvent.
1. A laminated retarder stack comprising
a first organic layer, the molecules of the first organic layer substantially having a first orientation; and
a second organic layer chemically bonded to the first organic layer, the molecules of the second organic layer substantially having a second orientation;
wherein the first and second orientations of the chemically bonded first and second organic layers are crossed.
2. A laminated retarder stack according to
3. A laminated retarder stack according to
4. A laminated retarder stack according to
5. A laminated retarder stack according to
6. A laminated retarder stack according to
7. A laminated retarder stack according to
8. A laminated retarder stack according to
9. A laminated retarder stack according to
10. A laminated retarder stack according to
11. A laminated retarder stack according to
12. A laminated retarder stack according to
13. A laminated retarder stack according to
14. A laminated retarder stack according to
15. A method for laminating organic films into a retarder stack, the method comprising:
providing a first organic layer, the molecules of the first organic layer substantially having a first orientation;
providing a second organic layer, the molecules of the second organic layer substantially having a second orientation; and
bonding the first organic layer to the second organic layer with a chemical solvent;
wherein the first and second orientations of the bonded first and second organic layers are crossed when the first and second organic layers are bonded together.
16. A method according to
17. A method according to
18. A method according to
19. A method according to
20. A method according to
This application is a continuation of U.S. application Ser. No. 10/970,029 filed on Oct. 22, 2004, which is a divisional of U.S. application Ser. No. 09/126,330 (U.S. Pat. No. 6,882,384) filed on Jul. 31, 1998. All of the above-listed applications are incorporated by reference in their entirety for all purposes.
The invention generally relates to laminated retarder films. More particularly, the invention relates to retarder films that are chemically-bonded to each other.
Pressure sensitive adhesive (PSA) typically has been used to bond of sheets of polycarbonates to form a retarder stack. Such retarder stack is comprised of several sheets of polycarbonate with acrylic based PSA's between them. An interface exists between each PSA and polycarbonate layer.
Transmission losses in retarder stacks are the result of several loss mechanisms. Losses are a consequence of material absorption and random scatter, as well as reflections at the interface between each PSA and polycarbonate layer. Reflection at the interface between each PSA and polycarbonate layer is typically dominant, as PSA material often has a refractive index far below that of the polycarbonate. Thus, methods for increasing transmission focus on eliminating surface reflections.
The present disclosure relates to retarder stacks with minimized surface reflection. In one embodiment of the disclosed retarder stack, a first organic film having a first molecular orientation is chemically bonded to a second organic film having a second molecular orientation. When the organic films are bonded, the first and second molecular orientations are crossed. The organic films may be a biaxially or uniaxially stretched film. In some embodiments, the organic films are made from polycarbonate polymer film. In some other embodiments, the organic films may be a polymer selected from the group consisting of poly-vinyl alcohol, polysulphone, polyacrylate, mylar, polyamide, polypropylene, polyethylene teraphalate, triacetate, polytetrafluoroethylene, and polymethylmethacrylate. In an exemplary embodiment, the organic films are welded together using a suitable solvent. An exemplary suitable solvent for laminating polycarbonate films is methylene chloride. In addition to the first and second organic films, the retarder stack may include additional films that are chemically bonded to at least one of the first and second organic films. Related methods for preparing the chemically-bonded retarder stack are also described.
The invention will be described in detail with reference to the following drawings in which like reference numerals refer to like elements wherein:
Color selective polarization modulators are described in the present invention for use as color selective light modulators (CSLMs). CLSMs are units that color selectively control the transmission of light using active or spatially patterned elements. The color selective elements in the CSLMs are typically stacks of retarder films. However, they can be thin twisted liquid crystal films. Two retarder stacks, which are related in retardation values and orientations, are used to determine the spectral band over which the active element can modulate light transmission. An polarization modulator or a spatially patterned polarization modulator is positioned between the stacks for controlling the transmission of selected spectral bands. If an active modulator is used, the transmission of the selected bands are electronically and temporally controlled. If a spatially patterned modulator is used, the transmission of selected spectral bands are spatially controlled.
Such CSLM devices are useful for common-path color display systems, digital color sequencers, and for analog color generation in applications such as lighting, spectrometry, colorimetry, and multi-spectral imagery.
A few definitions that are used to characterize the structures of the present invention and their behavior will now be provided. An additive primary color is one that is identifiable as a substantially saturated red (R), green (G) or blue (B) by a viewer with normal color vision. Conversely, colors that are desaturated RGB hues are termed pastels. Colors that are saturated but have other hues can be produced as admixtures of the additive primary colors, i.e., yellow or orange.
The term “subtractive primary color” refers to the absence of one additive primary color. These are known as cyan (C), magenta (M) and yellow (Y), and are the colors used in subtractive systems, i.e., printing systems. Unless otherwise stated, cyan is considered to be blue+green, magenta is considered to be blue+red, and yellow is considered to be green+red. Monochromatic yellow is referred to as true-yellow.
Colors are complementary when the sum of their power spectra gives a true white spectrum, i.e., cyan+red, magenta+green, or yellow+blue. By conservation of power, complementary colors can be generated, for instance, by rotating a neutral analyzing polarizer, following an arbitrary polarizer/retarder stack, through (π/2).
A “filtered” spectrum is a spectrum that is chromatically altered upon transmission through the filter structure. The transmitted spectrum has a particular hue, and is significantly more saturated than a “neutral” spectrum discussed below. Preferred filtered spectra are those that represent saturated colors, such as the additive primaries RGB, or the subtractive primaries CMY.
A “neutral” spectrum is one that is not chromatically altered by upon transmission through the filter structure. The transmitted neutral spectrum has coordinates that are substantially nearer the white point of the Commission International DE l'Eclairge diagram than a filtered spectrum. Examples include the white state or the black state of a filter.
An “Eigenpolarization” is a polarization state which is unchanged upon passage through a filter structure.
“Stack-Only” mode refers to the output of a filter structure when the active element is absent, or when the active element is driven to the electrooptic equivalent of being absent. The SO mode can be such that the two retarder stacks work cooperatively to transmit either a neutral spectrum or a filtered spectrum when they are positioned between neutral polarizers.
“Stack-Altered” (SA) Mode refers to the output of a filter structure when the active element is used to effectively alter the “presence” of one stack relative to the other stack. The SA mode can be such that the two stacks work cooperatively to generate either a neutral spectrum or a filtered spectrum when they are positioned between neutral polarizers.
The human eye has a wavelength-dependent sensitivity, being most sensitive to green light during normal daylight illumination conditions. Conversely, the eye is nearly insensitive to short wavelength blue, i.e., less than 400 nm, and long wavelength red light, i.e., greater than 700 nm.
The eye has three discrete photoreceptors that correspond to the three additive primaries. As such, the eye can be “tricked” into observing arbitrary colors using appropriate mixtures of the three additive primaries. This allows for the generation of full-color imagery using only three-color information. The additive primaries are separated by transition bands that are, in principle, not necessary for full color imagery. These transition bands are true cyan, at a wavelength of approximately 500 nm, and true yellow, at a wavelength of approximately 585 nm. In many instances, the transition bands are rejected in order to enlarge the available color space.
The CSLM structures of the present invention comprise two stacks of retarders, with an active polarization modulator or a spatially patterned polarization modulator between them. The polarization modulator can preferably be driven between two extreme states, preferably in an analog fashion. In a preferred embodiment, the modulator enables the SO mode in one extreme state, and enables the SA mode in the other extreme state.
If a spatially patterned modulator is used between the two retarder stacks, the spatially patterned modulator is preferably patterned so that the SO mode is enabled when light passes through one portion of the patterned modulator, and the SA mode is enabled when light passes through another portion of the patterned modulator. In addition, one or more additional portions of the modulator may be patterned to enable additional intermediate modes, i.e., modes that lie in between the SO mode and the SA mode.
In the SA mode, the output observed between neutral polarizers is the result of the cooperative effects of the retarder stacks, in conjunction with the effect of the modulator. In effect, the modulator modifies the presence of one retarder stack relative to the other, thereby modifying the characteristics of the both retarder stacks together. The SA mode is associated with an extreme modification of the both retarder stacks relative to the SO output. However, it should be recognized that analog states can exist between the SO and SA modes. The above discussion applies to all the filter structures described herein.
The methodology for designing individual retarder stacks, as well as the relationships between retarder stacks, depends upon the characteristics of the polarization modulator (modulator) used in the CSLM. Modulators can include, but one not limited to, mechanical, electromechanical, liquid crystal, linear electrooptic effect, magneto-optic, and nonlinear electrooptic effect devices. Modulators can be addressed electrically, optically, or using hybrid effects via media that generate photo currents.
Liquid crystal (LC) based modulators are preferred due to their low manufacturing costs, their scalability, their ability to impart large polarization transformations per unit interaction length, their ability to impart large modulation with low voltage/power, and other considerations. LC devices generally fall under the following categories: (1) Zero-Twist Nematic (ZTN) or Electrically Controlled Birefringence (ECB) devices; (2) Twisted Nematic Devices; and (3) In-Plane Switching Devices.
ZTN or ECB devices are a subset of LC devices in which the in-plane retardation is controlled by an electric field applied normal to the plane. ZTN or ECB devices include anti-parallel nematic devices, pi-cell or surface mode devices, vertically aligned nematic (VAN) devices, and hybrid aligned nematic (HAN) devices.
Anti-parallel nematic devices and pi-cell or surface mode devices have maximum retardation in the undriven state and are nearly isotropic in the driven state. High contrast is often obtained by adding external compensation in order to eliminate residual retardation in the driven state.
VAN devices are homeotypcially aligned on both substrate surfaces. They are isotropic in the undriven state, with an increasing in-plane retardation with applied voltage. VAN devices have very little residual retardation and hence give high contrast with no external compensation.
HAN devices are homeotypically aligned on one substrate surface, and planar aligned on the other substrate surface.
A Twisted Nematic device is one in which the substrates are rubbed other than parallel or antiparallel. Twisted Nematic devices with a difference in rub angle of approximately 45°-90° are typically called “TN” devices, while devices with a difference in rub angle of approximately 240°-270° are typically called “STN” super-twist nematic devices. TN devices are typically addressed using transmissive active matrix thin film-transistor (TFT) panels. STN devices can be highly multiplexed, permitting passive matrix addressing schemes. Small twist angles (40°-60°) are often used for mixed-mode reflective displays.
An In-Plane Switching device refers to any switch in which the molecular directors rotate substantially in the substrate plane. Examples include chiral smectic devices, e.g., ferroelectric LC or electroclinic LC devices, surface-stabilized chiral smectic C, volume-stabilized smectic C, distorted helix ferroelectric, flexoelectric, achiral smectic, and anti-ferroelectric LC devices. There are also In-Plane Switching nematic devices which are driven by lateral electrodes. That is, if lateral electrodes on one of the device substrates are used rather than transparent electrodes deposited on opposing device substrates, nematic liquid crystals can operate as rotatable retarders with fixed retardance.
Compound retarders utilizing a liquid crystal active retarder in combination with one or more passive retarders can be used as the modulator in the CSLM. Particularly useful are achromatic compound retarders, as described in U.S. patent application Ser. No. 08/419,593, and achromatic polarization rotators, as described in U.S. patent application Ser. No. 08/549,963, both of which are incorporated by reference in their entirety herein.
The achromatic compound retarder comprises a liquid crystal rotatable half-wave retarder flanked by passive retarders, wherein the orientations and retardances of the passive retarders are such that the compound retarder is achromatic. The achromatic polarization rotator comprises a liquid crystal rotatable half-wave retarder in combination with a passive half-wave retarder. In the optically compensated bend (OCB) cell, a pi-cell is combined with a retarder film that has a z-direction retardence to improve field-of-view. STN displays frequently use one or more retarder films to suppress birefringence color that is otherwise observed. Both TN and STN displays use retarder films to improve contrast ratio and to improve the view angle.
Typically, the behavior of LC devices is interferometric in nature. Specifically, there are two orthogonally polarized eigenpolarizations with a voltage-controlled phase shift between them. The phase difference introduced between waves, i.e., the retardation, results in a change in the state of polarization (SOP). ZTN devices are very simple, because the eigenpolarizations are well defined and are both independent of voltage and wavelength. Though more complex, twisted nematic devices also have eigenpolarizations. The specific designs presented herein are optimized for ZTN modulators, though it should be appreciated that retarder stacks can be designed and generated for use with twisted nematic devices.
Optically active devices are another group of electro-optic modulators that can be used as the modulator in the CSLMs of the present invention. Optically active devices are polarization rotators, with the polarization rotation being independent of the polarization orientation of the incident light. The isotropic states are left-handed and right-handed circular polarizations. The optically active modulator performs voltage-controlled phase shifting of the circular polarizations, which does not affect the circular state of polarization. However, for linear states, which can be decomposed into equal amplitudes of the left-handed and right-handed circular polarizations, the transmitted polarization is linear and the orientation is determined by the phase shift between the circular states. Thus, optically active modulators act as polarization rotators for linear states, with the modulation state of polarization being any linear state.
Chiral Smectic Liquid Crystal (CSLC) retarders reflect polarization about an axis rather than rotating the polarization. However, they are like optically active devices in their isotropic and modulation states of polarization. In contrast to zero-twist nematics, CSLCs rotate in-plane, having a fixed retardence and an optic axis orientation that is determined by the applied voltage. For a half-wave retardence, CSLC retarders have circular isotropic states and linear modulation states. The voltage-controlled orientation of the optic axis in the CSLC retarder determines the transmitted orientation of linear light.
There are also electro-optic modulators that have intermediate modulation and isotropic states, such as twisted nematic devices. The polarization transformation of a twisted nematic device is most easily seen on the Poincare sphere (J. E. Bigelow and R. A. Kashnow). The State of Polarization follows a cone rolling on the equator, with a base radius that depends, in part, on wavelength. As the pitch, which is four times the cell gap for 90-degree TN, and birefringence become large, the base of the cone becomes small and almost independent of wavelength, and the device appears optically active, i.e., it exhibits circular Eigenpolarizations. For lower pitch values, a linear state introduced along the rub direction, i.e., normal direction, performs excursions from the equator. These excursions represent periodic introduction of ellipticity coupled with the dominant rotation. Light introduced along the rub direction with the appropriate ellipticity maintains that ellipticity during rotation. This wavelength-dependent elliptical state experiences a pure rotation. This construction can also be used to describe the transformation of an STN device in the undriven state.
It was recently shown experimentally that a twisted nematic device is an elliptical retarder with elliptical Eigenpolarization states. Like the linear variable retarder, the phase shift is substantially determined by one particular helicity of polarization state. This is described in J. L. Pezzaniti and R. A. Chipman, “Phase-only modulation of a twisted nematic liquid-crystal TV by use of the eigenpolarization states,” Opt. Lett. 18, 1567-1569 (1993).
To achieve compatibility between the retarder stacks and the modulator, passive retarders can be included in the CSLM. For example, if the retarder stacks prepare the light in two linear polarizations separated by 45°, these polarization states are the modulation and isotropic states of polarization for a modulator that comprises a variable retarder with fixed orientation. A passive quarter-wave retarder oriented parallel to one of the polarization states converts the two linear polarizations into one linear and one circular polarization, which are the modulation and isotropic states of polarization for a modulator that comprises a rotatable retarder with fixed retardence. Thus, if the CSLM includes a passive quarter-wave retarder on either side of a modulator that comprises a CSLC retarder, retarder stacks designed for a variable retarder can instead be used with a rotatable retarder.
Without using specific retarder stack designs, an analysis of filter designs can be performed subject to the following assumptions: (1) it is desirable to generate either black or white in the neutral state of a color filter/switch; (2) it is desirable to generate “stable” primary colors (nominally from the set RGBCMY) in the filtered state of the color filter/switch. That is, it is desirable for the chrominance of the modulator, e.g., an LC cell, to have little influence on the filter color coordinate; (3) the CSLM is designed so that the modulator optically “vanishes” in the SO mode throughout the visible spectrum; and; (4) the modulator exhibits some chrominance in the SA mode, such that it cannot adequately maintain a constant polarization transformation throughout approximately two-thirds of the visible spectrum. With those assumptions, various general color filter configurations can be analyzed. It should be noted that there may be applications for which the above assumptions are not relevant.
If an LC cell is used as the modulator one of the most significant differentiators when comparing color filter/switch designs is the impact of the LC cell chrominance on the quality of the output. The SO mode is preferably selected to be the output that can least afford to suffer the negative effects of cell chrominance. The ultimate decision regarding the SO and SA modes is preferably based on a global view of overall component performance, which may include multiple color filter/switch stages and multiple polarizers. The priorities are derived from the specifications for a particular application, these include, but are not limited to, the number of outputs, the location of center wavelengths, the bandpass profile, and the throughput requirement for each band. For example, a color sequencer for a digital camera may have optimum spectra that are quite poor for display applications. Other practical issues can also come into play, including the number of allowed polarizers, limitations in drive voltage, characteristics of lamp spectra, and field-of-view requirements. There is clearly no “best-design” that will cover all applications. However, a few examples will be described to illustrate how the present invention can be adapted for different applications.
Most applications will utilize multi-stage color filters, with each color filter stage preferably comprising a pair of retarder stacks with a central LC cell as the modulator. In a multi-stage design, it is preferable for the color filter stages to function independent of one another, so that they can be independently designed and be completely modular. This simplifies the design process, particularly in multi-stage configurations. For example, the retardence of one retarder stack in a multi-stage color filter may be shifted to accommodate the needs of one application, with no impact on the outputs of the remaining color filter stages.
In the interest of maintaining the independence of color filter stages, and obtaining a high density black state or an extremely neutral white state, the SO mode preferably corresponds to the neutral output of each color filter stage. In some instances, however, the designer of the color filter stages does not have this option. For example, in order to have independent color filter stages, each color filter stage must generate an acceptable filtered state, i.e., color coordinate, with no assistance from other color filter stages. Even when an additive primary color is obtained subtractively from two color filter stages, a lack of redundancy in color blocking remains. In other words, an acceptable additive primary color coordinate cannot be achieved unless each color filter stage can generate a high quality subtractive primary color. As a result, priority is frequently given to obtaining acceptable filtered states over ideal neutral states.
There are design schemes that provide a high quality SA filtered state, using chromatic switches, thereby preserving the SO mode for the neutral state. In these design schemes, the LC cell need only switch one additive primary color. Examples of this include R/K (red/black), G/K and B/K, as well as the complements C/W, M/W and Y/W. Assuming that the LC cell is more chromatic in the SA mode than in the SO mode, the density of the black state (or flatness of the white state) is better using SO black (white). Typically, the primary color quality is indistinguishable for SO or SA filtered modes. Stacks of these types of color filter stages are desirable for multi-stage color filters.
When the LC cell must switch between two or more additive primary colors, the situation changes. SA additive primary with SO white suffers significant degradation in the filtered color coordinate. This is particularly so for SA green, because the active element must rotate the polarization of both blue and red in order to effectively block them. This requires an LC cell that is achromatic throughout the visible spectrum. Accordingly, the SO mode is devoted to the filtered state in order to preserve the color coordinate. Consequently, the white state can have significant short-wavelength blue and long-wavelength red loss. The white state actually appears as if the retarder stacks were removed, leaving only the LC cell oriented at 45-degrees. The retarder stacks, in many cases, improve white transmission, but never appear to degrade transmission when compared to the LC cell alone.
It is assumed that at least partially polarized light is input to the CSLM. While unpolarized light can be introduced, it should be understood that randomly polarized light is not manipulated by the CSLM, and is passed through the CSLM unmodulated. Therefore, we consider the action of the CSLM on the polarized component of the input light. Furthermore, it is assumed that the polarized component of the input light is neutrally polarized over a particular operating band. A “white” state is thus defined as an output transmission that is substantially neutral throughout this particular operating band.
An optimum white state is defined as a transmission function with unity transmission throughout the operating spectral band. For visible operation, the operating spectral bond is 400-700 nm. Conversely, a filtered state is one in which the transmission function contains one or more maximum transmission points (ideally zero loss), and one or more minimum transmission points (ideally zero transmission) within the operating band.
The requirements placed on a single color filter stage for achieving a neutral state will now be discussed. The discussion will focus, in particular, on a maximum throughput white state, with the understanding that such a design can also represent a black state by a suitable rotation of the analyzing polarizer.
Polarized input light can be generated directly, e.g., from a laser, or can be generated by placing a passive polarizer between the light source and the CSLM. While many types of passive polarizers exist, the most common passive polarizers generate linear polarization states. Therefore, the discussion and examples presented concentrate on the special case of linearly polarized input light. However, it should be understood that other neutral light inputs can be generated directly or indirectly using an elliptical polarizer, or an achromatic retarder structure in combination with a neutral linear polarizer.
The color filter stage comprises a retarder structure that contains first and second retarder stacks on either side of an active polarization modulator and an analyzing neutral polarizer (analyzer). The term “neutral” signifies that the polarizer provides substantially constant transmission and polarizing efficiency throughout the operating band. The operating band contains at least one spectral portion that is substantially modulated by the modulator and at least one spectral portion that is substantially unmodulated by the modulator.
If the input light polarization differs from the polarization that is transmitted by the analyzer, an efficient white state requires that the retarder structure perform a neutral polarization transformation. This transformation can be accomplished in either the SO or SA modes. For the SO mode, the retarder stacks cooperate to produce a neutral transformation that matches the input light polarization to the exit polarization. The modulator, e.g., the LC cell, thus switches the CSLM structure to one that is substantially more chromatic, thereby accomplishing the desired filtering operation.
For the SA mode, a neutral transformation is accomplished via the action of the retarder stacks with the modulator, or in some cases by the modulator alone. An example of this is the specific case of a linear input polarization state and a linear analyzing polarizer. In the neutral state, using either SO or SA modes, the modulator must rotate the polarization by θ for a highly efficient white state, or by θ+π/2 for a black state, where θ is the orientation of the analyzing (exit) polarizer. Because the modulator performs a relatively neutral operation, it is attractive to consider an arrangement in which the modulator is largely responsible for the characteristics of the neutral state. While the modulator may not be particularly achromatic, an analog modulator may be tuned to position the peak transmission of the neutral spectrum to maximize performance. For instance, a multi-stage color filter may maximize throughput by tuning the peak transmission of the color filter stages in the neutral state to coincide with the peak transmission of a color filter stage in the filtered state.
In the specific case where the analyzing polarizer analyzes the input state of polarization, e.g., x-oriented input polarization and x-oriented linear analyzing polarizer an optimum white state is obtained when the retarder stack structure provides a zero net polarization transformation. This is equivalent to requiring that the input/output polarization be an eigenpolarization of the retarder stack. In this case, a preferred subset of retarder stack designs is one in which the input/output polarization coincides with an eigenpolarization of the retarder structure for one particular state of the LC device. This includes cases where additional retarder stacks are used to produce a neutral polarization state before the chromatic retarder stack, and analyze the neutral polarization state after the chromatic retarder stack, i.e., a Pancharatnam achromatic retarder preceding and following a linear polarizer. However, such solutions are more complex because they require additional retarders.
In the case in which matched input/output polarizations produce a white state in the SO mode, the first retarder stack performs a chromatic polarization transformation that must be fully compensated by the second retarder stack within the operating band. This simply requires that the input polarization correspond to an eigenpolarization of the combined retarder stack. For example, a zero-oriented linear polarization state with an input retarder stack having individual retarder orientations of −7.5°, 52.5° and 67.5° can be followed by the second retarder having individual retarder oriented at stack 67.5°, 52.5° and −7.5°, yielding a neutral white state. Note that this particular structure is symmetric in both retardation and orientation, thus enabling true reflection mode color filters/switches. It should be further noted that this retarder stack structure requires a specific input State of Polarization, and does not perform identically for other input polarization orientations, e.g., 45° linear or circular input polarizations.
A “crossed-stack” arrangement provides a family of solutions for SO white that does not depend upon a specific input/output polarization. When two retarders have equal retardation and have crossed orientations, the resulting Jones matrix is the identity matrix. This concept can be extended by adding additional retarders on either side, with each additional pair having the same retardation and crossed orientations. For any number of retarders conforming to this scheme, there is zero net polarization transformation because the Jones matrices representing each retarder stack are inverses of one another. Because the retarder structure appears isotropic, an optimum white state is obtained for any matched input/output polarization. Optimum stack designs for filtered outputs in the SA mode depend on behavior of the modulator and the chromatic polarization states that are optimally presented by the input retarder stack.
Another case is that in which matched input/output polarizations produce a white state in the SA mode. In this case, the input polarization corresponds to an eigenpolarization of a CSLM structure that now includes both the retarder stacks and the modulator. For example, a zero-oriented linear polarization state with an input retarder stack having individual retarders oriented at −7.5 °, 52.5° and 67.5°, can be followed by a second retarder stack having individual retarders oriented at −67.5°, −52.5° and 7.5°. This yields a white state with a ZTN half-wave plate (modulator) oriented parallel to the input. The structure thus appears to have the eigenpolarization of the symmetric structure described above.
In a second example, the modulator can be used to create the appearance of crossed retarder stacks. Using the above-described input polarization, input retarder stack, and ZTN half-wave plate orientation, the second retarder stack has individual retarders oriented at 22.5°, 37.5° and −82.5°.
The demands placed on the color filter structure from a transmission spectrum perspective will now be discussed. As discussed previously, the operating band is defined as having at least one spectral portion in which the polarization is substantially unmodulated by the modulator, e.g. LC device, and at least one spectral portion in which the polarization is substantially modulated by the modulator. In a preferred embodiment, the polarization modulation corresponds to a transmission modulation by a suitable selection of an analyzing polarizer. Furthermore, the modulated spectral portion is preferably modulated between substantially full transmission and substantially zero transmission. This requires that the modulator be capable of modulating between orthogonal polarization states within the modulated portion of the spectrum. The unmodulated portion of the spectrum must retain its polarization state independent of the voltage applied to the modulator. Thus, the transmission must remain substantially fixed (typically full transmission or zero transmission) for any voltage state.
The above design considerations are considered with respect to the following filter structure: a modulator positioned between two retarder stacks. The filtered portion of the spectrum is fully modulated between orthogonal states and the unmodulated spectral portion does not “see” the modulator in a way that is significant to its transmission through the filter structure. The polarization of the filtered portion of the spectrum must be properly configured for modulation as in a conventional neutral shutter. The unmodulated portion corresponds to a zero-modulation state of the modulator, i.e., a modulator eigenpolarization.
The modulator, whether single-layer or compound (i.e., multi-layers of active and/or passive elements), twisted or zero-twist is preferably a device that is interferometric in behavior. That is, there are two orthogonal eigenpolarizations, each with an optically distinct path length. It is assumed that the path length between these waves can be controlled. It is the path length difference between waves, that project onto each eigenpolarization, that influences the state of polarization of light transmitted through the modulator. Preferred modulators for the CSLM structures of the present invention are those that have voltage and wavelength stable eigenpolarizations, such as zero-twist nematics (ZTN). Conversely, twisted devices have eigenpolarizations that are relatively sensitive to wavelength and voltage. While many of the structures of the present invention are based on a wavelength effect that is stable over the modulated band, the chrominance of the modulator can, in some instances, be used to enhance the color selection technique. For example, the retarder stacks can open multiple spectral “windows” over which a modulator can modulate. A chromatic modulator can then be used to select which of the spectral bands is transmitted and which is blocked. Accordingly, the performance of most of the structures of the present invention is enhanced by suppressing modulator chrominance.
In view of the above discussion, there are exemplary input polarizations for achieving minimum modulation and for maximum modulation potential. Minimum modulation polarizations have projections only along a single eigenpolarization. For a ZTN device, for instance, the minimum modulation polarizations are the ordinary and extraordinary waves. Maximum modulation polarizations correspond to polarization states with substantially equal projections along the eigenpolarizations. For a ZTN device, this is the family of polarizations with arbitrary ellipticity, but with an ellipse axis oriented at ±π/4 to the rub direction. Thus, a retarder stack can generate, for example, linear ±π/4 polarizations or circular polarizations in a particular band, and the modulator can provide full modulation capability for each polarization state.
Evaluating the quality of the filtered state is somewhat complex because since no single design is “best” for all applications and output colors. Design considerations such as the width of the passband relative to the stop-band, and the transition slope, all depend upon the specific color being filtered and the application requirements.
Preferred spectra for display applications have a frequency domain duty ratio appropriate to the desired spectrum, and have a spectral profile that is rectangular, rather than sinusoidal or some other spectral profile. There may be additional physiological or other reasons why other spectral profiles are preferred. In general, however, the filter should probably have a steep transition slope, with a relatively flat transmission band and a relatively flat blocking band. Relatively flat bands can be achieved by having a series of distributed high-contrast nulls and peaks in the spectrum. Near-ideal transmission spectra can be produced by increasing the number of retarders in the retarder stacks. In practice, acceptable transition slopes and side-lobe amplitudes/locations must be carefully chosen to optimize saturation with a limited number of components.
The criteria for evaluating filter designs is based on considerations of saturation, hue, and throughput. The saturation and hue can be evaluated using the CIE chromaticity diagram. The quality of color generated by a particular filter output can be characterized by calculating a series of overlap integrals including the transmission function for a specific filter state the power spectrum of the source, and the CIE color matching functions.
Saturated primary colors are generated by maximizing the ratio between the light source power transmitted in the desired primary band to the light source power transmitted outside of the primary band. The filter design can be matched to the light source characteristics to make the optimization very specific. For example, true white sources, such as a 6000 K black body, place greater demand on filter performance than distributed light sources, such as a CRT phosphor or a fluorescent backlight. The spectral positions of nulls in the blocking band depend upon the retardence of the optical components. It is advantageous to strategically place nulls at out-of-band power spectral maxima of the light source. Similarly, it is advantageous to place side-lobe maxima away from out-of-band power spectral maxima. Passive filters can be inserted to reject bands that lie outside of the primary color bands to increase saturation.
The polarization transformation accomplished by each retarder stack is arbitrary, being determined by the retardation and orientation of each retarder film. However, constraints can be placed on retarder stack designs in order to obtain particular results. For instance, when a zero-twist nematic LCD is used as the modulator, the polarization of the center wavelength can be constrained to be a ±π/4-oriented linear polarization exiting the retarder stack. Consider a retarder stack that is comprised of equal retardation elements, each with a retardation element providing a half-wave retardence, of arbitrary order, at the center wavelength. The following equation must be satisfied:
where m is an integer, N is the number of retarder elements, and the final sign is determined by whether N is odd or even.
There are also designs in which the peak modulated wavelength is not a ±π/4-oriented linear state after the first retarder stack. Consider a red color filter that conforms to a quasi folded Solc design. The retarder stack is made up of polycarbonate retarders which are full-wave at 450 nm. In this case, with N=3, the rocking angle should be 7.5° for a π/4 rotation angle. However, the red color filter design uses a significantly larger 12° rocking angle, giving a 72° rotation angle. This rotation angle corresponds to an 807 nm half-wave wavelength. Not surprisingly, the efficiency is poor for this wavelength, though it lies well beyond the visible. Nevertheless, the efficiency is high at the much shorter 668 nm wavelength, where the retardance is significantly larger than half-wave. In this particular case, the larger rocking angle compensates for the loss in rotation efficiency at the shorter wavelength, yielding the desired intermediate polarization at 668 nm. Full red transmission is possible because the retarder stack generates a ±π/4-oriented elliptical state.
In U.S. patent application Ser. No. 08/645,580, which is herein incorporated by reference in its entirety, structures were disclosed that generate an achromatic white state or black state by producing an identity Jones matrix between parallel polarizers for the white state, or crossed polarizers for the black state. For an arbitrary input retarder stack, the inverse retarder stack is produced by reversing the order of the individual retarders, and rotating each individual retarder π/2 from the corresponding individual retarder in the input retarder stack. The transmission is unity for all wavelengths, assuming lossless components and assuming that the modulator between the crossed retarder stacks is made to vanish. In the crossed retarder stack arrangement, the white state quality is insured regardless of the modulator orientation or the electrooptic effect, provided the modulator can be made to effectively vanish. Thus, the filtered state can be designed without having to consider the impact on the white state quality. One result of such independent design characteristics is that SO neutral configurations can be readily adapted into multiple stage filter designs.
Part of the present invention is the discovery that, by orienting a zero-twist nematic (ZTN) LC modulator at 0 or π/2 between crossed retarder stacks, many useful filtering operations can be obtained. Assuming that the input retarder stack has individual retarders oriented at α1, α2, . . . αN, the second retarder stack has individual retarders oriented at π/2+αN, . . . π/ 2+α2, π/2+α1. Assuming that the ZTN LC modulator is a half-wave retarder over the modulated bandwidth, the ZTN LC modulator, in the SA mode, effectively switches the angle of each individual retarder in the second retarder stack to π/2−αN, . . . π/2+−α2, π/2+−α1.
Accordingly, the states are inverted by simply giving the individual retarders in the second retarder stack the effective orientation produced in the SA mode with crossed retarder stacks. This produces a SA white state, subject to the effects of the LC modulator chrominance on white-state transmission. Such SO filtered configurations are valuable when the effects of LC modulator chrominance on the quality of a filtered state are intolerable.
In addition, the SA white state quality is substantially determined by the LC modulator chrominance alone and, as such, can be tuned to provide maximum transmission at an arbitrary wavelength. This is important for providing high throughput in multi-stage SA white filters. The effect of the retarder stack on the white state, if any effects are observed at all, is to increase the transmission. The retarder stack has never been found to degrade the transmission of the white state.
Designs in which the SA white states are compromised by the LC modulator chrominance alone are generally preferred over designs with SO white outputs with similar chrominance. SO white output designs refer to designs that use the retarder stacks to produce a quasi-neutral transformation to yield a white state. The SA white designs are preferred in multi-stage filters because the white state can be tuned to maximize throughput. In contrast, the SO white state has a fixed chrominance that represents a loss of throughput.
Examples of CSLM devices that represent a range of design options for modulating between neutral and filtered transmission spectra will now be discussed. All of the CSLM devices discussed use ZTN LC devices for the modulator. However, different options for the modulator orientation, retarder stack interactions, and polarizer orientations will be discussed.
One embodiment of the CSLMs of the present invention comprises crossed retarder stacks and parallel polarizers, and gives a substantially ideal SO white output. The CSLM structure is shown in
Because only one additive primary band is switched by the ZTN LC modulator, the effects of modulator chrominance on the SA spectrum are almost undetectable. The white spectrum is substantially ideal because the first and second retarder stacks 110 and 120 are crossed.
The first retarder stack 110 comprises individual retarder films 210 a, 210 b and 210 c oriented at 7.5°, −7.5°, and 7.5°, respectively, with respect to the input polarizer 130. The second retarder stack 120 comprises individual retarder films 220 a, 220 b and 220 c oriented at 82.5°, −82.5°, and 82.5°, respectively, with respect to the input polarizer 130. The individual retarder films in the first and second retarder stacks 110 and 120 are 1.5-wave retarders in the grain (140 nanometers). Accordingly, the first and second retarder stacks each give a π/4 polarization rotation (half-wave retardence) for green light. The exit polarizer 140 is oriented with its polarization axis parallel to the polarization axis of the input polarizer 130. Accordingly, the light output is modulated between a magenta state and a white state.
For orientations of the ZTN LC modulator 100 other than 0 or π/2, designs can be generated for SA white states by considering the effective input to the second retarder stack 120. For example, the orientation of the second retarder stack 120 and the exit polarizer 140 can be selected to provide effectively crossed retarder stacks and parallel polarizers. This white state is substantially ideal insofar as the modulator 100 is achromatic over the entire modulated band.
A modulator with orientation αLC produces an effective input polarization of 2αLC, with an effective input stack with orientations 2αLC−, α1, 2αLC−, α2, 2αLC−, αN. Thus, an effective crossed stack white output can be generated by orienting the exit polarizer 140 at 2αLC and the second retarder stack at π/2+2αLC−αN, . . . π/2+2αLC−α2, π/2+2αLC−α1.
In order to obtain a M/W output like the color filter shown in
The first retarder stack 110 comprises individual retarders 310 a, 310 b and 310 c, oriented at 22.5 °, 75° and −60°, with respect to the input polarizer 130. The second retarder stack 120 comprises individual retarders 320 a, 320 b and 320 c, oriented at 30°, −15° and −67.5°, respectively, with respect to input polarizer 130.
The embodiment shown in
An alternative to crossed retarder stacks is to achieve a neutral output using an eigenpolarization of the CSLM structure.
The CSLM 10 of
In the specific embodiment of
In the CSLM of
In CSLM 10 shown in
The configuration shown in
The CSLM embodiment shown in
A SA white state can be obtained as an eigenpolarization by reversing the signs of the angles of the individual retarders in one of the retarder stacks (110 or 120) of the CSLM 10 shown in
A CSLM embodiment that incorporates disconfiguration is shown in
When the chrominance of the modulator is too large for generating spectra with adequate saturation in the SA mode, such spectra are best generated in the SO mode. Consequently, the white state reflects the loss in transmission characteristic of most LC shutter devices in the open state. However, like ZTN neutral shutters, the peak transmission of the white state of the CSLM can be tuned to maximize throughput.
The ZTN LC modulator 100 is tunable so that the peak transmission of the white state of the CSLM 10 can be tuned to maximize throughput. The input and output polarizers 130 and 140 have crossed orientations. The first retarder stack 110 comprises individual retarders 910 a, 910 b and 910 c oriented at 7.5°, −7.5° and 52.5°, respectively. The second retarder stack 120 comprises individual retarders 920 a, 920 b and 920 c oriented at −52.5°, 7.5° and −7.5°. The ZTN LC modulator 100 is oriented at 45°.
The first retarder stack 110 comprises a plurality of individual retarders 1010 a, 1010 b, . . . 1010 c having retardances Γ1, Γ2, . . . ΓN, respectively, and orientations α1, α2, . . . αN, respectively. The second retarder stack 120 comprises a plurality of individual retarders 1020 a, . . . 1020 b and 1020 c having orientations α′N, . . . α′2, α′1. The individual retarders and the second retarder stack 120 have the same retardances as the individual retarders in the first retarder stack 110, but with the order reversed. It should be appreciated that, for illustrative purposes, each of the first and second retarder stacks 110 and 120 are shown with three individual retarders. It should be understood that the number of individual retarders included in each of the retarder stacks 110 and 120 can be less than or greater than three.
A reflection mode CSLM 20 is shown in
To more clearly illustrate the path light takes through the reflection mode CSLM 20, an “unfolded” version of the reflection mode CSLM is shown in
The reflection mode CSLM 20 has the constraint that α′i=ai. While achromatic transformations can be inserted to broaden the range of design choices (Ammann and Yarborough), it is most desirable to avoid such transformations, thereby simplifying the stack design. In a retroreflecting arrangement, the unfolded version shows that the polarizers are parallel and the layers in corresponding retarder stacks have identical orientations. If a polarization splitter is used, there is an option between a parallel polarizer output or a crossed polarizer output.
The CSLM embodiments discussed above primarily utilize an LC modulator to control the transmission of one portion (spectrum) of the operating band, while maintaining fixed transmission of the complementary spectrum. Preferred embodiments are those in which the unmodulated spectrum is either fully transmitted or fully extinguished. The embodiments in which the unmodulated spectrum is fully transmitted represent variable saturation CSLMs, because the LC modulator modulates between a white state and a filtered state that is substantially more saturated. The embodiments in which the unmodulated spectrum is fully extinguished are variable luminance CSLMs, because the LC modulator modulates between a filtered state and a black state. In many instances, a variable luminance CSLM permits full modulation from a filtered state to a black state with fixed color coordinates.
Another type of filter is a variable hue filter. A variable hue filter is one with both end points on the CIE diagram corresponding to filtered states. The action of the LC modulator(s) is thus to modulate between two filtering operations, rather than a filtered state and a neutral state. Examples of some variable hue filters are provided in U.S. Pat. No. 5,751,384, the embodiments of which are incorporated herein by reference. CSLM structures also permit such operations.
In the CSLM embodiment of
When the IPS modulator 1310 is oriented at 5π/12, a green output with a 40 nm bandpass width is transmitted. When the IPS modulator 1300 is rotated to 2π/3, a magenta with half maximum points at 494 nm and 596 nm is output. The spectra overlap at the 10% transmission points.
Multiple stage CSLMs comprise two or more retarder stack/modulator stages in series with one or more analyzing polarizers. Each retarder stack/modulator stage comprises a modulator with a retarder stack positioned on each side of the modulator. The polarizers are typically neutral, but polarizers designed to analyze a particular spectral component can separate retarder stack/modulator stages. Examples of two-polarizer and four-polarizer full color multiple stage CSLMs are used by way of illustrating issues that arise when retarder stack/modulator stages are cascaded. Multi-stage CSLMs can be used as color sequencers, as will be explained in more detail below.
Note that the actual number of stages is arbitrary. In addition, although red, green and blue stages are shown for illustrative purposes in the CSLMs of
An important benefit of a two-polarizer multi-stage CSLM, such as the general one shown in
The challenge in designing a full-color multi-stage CSLM with one analyzing polarizer is to minimize the interaction between the retarder stack/modulator stages. Because the state of polarization is not analyzed after each retarder stack/modulator stage, the stages interact on a field basis, rather than on a power transmission basis. The two-polarizer design thus requires greater independence of stages than the four-polarizer design. Since a polarization analyzer does not isolate the stages, any polarization effects introduced by modulator chrominance (or off-normal input light) in the neutral state represents a degradation in color coordinates for any retarder stack/modulator stage in the filter state. As the quality of the neutral state improves (the density of the black state or the flatness of the white state), the independence of the stages likewise improves, making SO neutral states a desirable design choice.
A significant benefit of the two-polarizer design is the greater number of design options available relative to four-polarizer designs. In particular, the two-polarizer CSLM can be operated in an additive mode. As discussed above, stages behave as variable saturation filters in subtractive mode, and as variable luminance filters in additive mode. When only one polarization analyzer is used, designs that use a black state rather than a white state can be considered. A black state simply means that a particular stage “vanishes” between crossed polarizers, rather than parallel polarizers. The term “vanishes” is used to indicate that the polarization of input passing through the stage is preserved. Thus, it is as though the stage is not present. In the context of a multiple-stage CSLM, a black state does not preclude other output spectra, because other stages remain available to switch to the filter mode. Conversely, a four-polarizer filter, using all neutral polarizers, requires that white be a common output of every retarder stack/modulator stage. Thus, by simply rotating the analyzing polarizer, the CSLM is converted from a subtractive filter to an additive filter, with some important performance implications.
A benefit of the additive mode design is that it gives preference to the quality of the additive primary colors and black, rather that the subtractive primary colors and white, which is important in display applications. When the input and output polarizers are crossed, the quality of the black state is primarily compromised by any inability to make the structure between the polarizers vanish. Conversely, when the input and output polarizers are parallel, a black state relies on the structure between the polarizer to convert the polarization to the orthogonal state. Using a full-color subtractive mode structure, this requires the cooperation of three retarders stack/modulator stages, each responsible for rotating the polarization of one additive primary color to the orthogonal state. This is typically a formidable task.
Other benefits are associated with the generation of color. The additive mode design generates an additive primary directly using a single stage, requiring no cooperation between stages. Each stage can thus be designed independently, without regard for the characteristics of the other stages, provided that the neutral states are of adequate quality. Therefore, with an additive mode CSLM, stages ideally “vanish” in the SO mode and each stage is responsible for independently determining the transmission of one additive primary color.
When generating subtractive primaries additively, losses in performance due to interaction between stages is frequently of lesser consequence than when generating additive primaries subtractively. It should be noted that, by conservation of power, the transmission spectra of additive and subtractive filters are inverses of one another. With this in mind, it is chromatically more tolerable to have a true yellow notch in the yellow spectrum than a yellow side lobe in the blue spectrum. Similarly, it is chromatically more tolerable to have a true cyan notch in the cyan spectrum than a true cyan side lobe in the red spectrum. Most importantly, true cyan and true yellow notches in the white output are less significant than true cyan and true yellow side lobes in the black state.
In a projection display, maintaining high light efficiency is a high priority. In addition, a digital color sequencer (DCS) must handle the high light level of the projector lamp. Accordingly, internal polarizers are preferably eliminated from the DCS design. In order to not decrease throughput, the DCS preferably does not create a bottleneck in view time, or duty ratio. Accordingly, the transient behavior of the CSLM must be considered. Fortunately, the field of view requirement is relaxed for to direct view systems, which is beneficial in two-polarizer designs. Typically, a 10°-15° half-cone angle is acceptable. Given these priorities, a preferred embodiment for a projector DCS will now be described. Note that, while much of the discussion centers around single-pixel color sequencers, this example also applies to stacked three-stage, pixilated color displays.
A preferred CSLM design for the projector DCS is a two-polarizer additive (SO black) design, with a SA red stage 1440, a SA green stage 1430 and a SA blue stage 1420, as shown in
The blue stage 1420 comprises a first stack 1470, a 450 nm half-wave LC modulator 1480 and a second retarder stack 1490. The green stage 1430 comprises a first retarder stack 1500, a 540 nm half-wave LC modulator 1510 and a second retarder stack 1520. The red stage 1440 comprises a first retarder stack 1530, a 650 nm half-wave LC modulator 1540 and a second retarder stack 1550. The retardances and orientations of the individual retarders that comprise retarder stacks 1470, 1500, 1520, 1530 and 1550 are listed in
The design of
Second, the spectra from each stage of the
Third, the LC modulators 1480, 1510 and 1540 each need only be achromatic over one additive primary color band, while complementary-based designs (assuming redundant blocking) require achromatic behavior over two additive primary color bands.
Fourth, the retardances of the LC modulator 1480, 1510 and 1540 have no significant influence on the color coordinate of the three-stage DCS, unlike complementary-based filters. This makes the DCS very robust, because the spectra are independent of the detailed characteristics of the LC modulators.
Fifth, the design of
Independent optimization of each retarder stack allows one to achieve near ideal color coordinates without the loss associated with additional notch filters. Such filters might otherwise be required in complementary-based designs to reject interprimary light. Achieving acceptable color coordinates depends upon the lamp spectrum (and chromatic losses in the system), the number of individual retarder films in the retarder stack, and the specific primary band modulated by the stage. As the center wavelength of the bandpass increases, it becomes necessary to increase the number of retarder films to maintain performance. In the B/K stage 1420 for example, only three of retarder films are used for each of the first and second retarder stacks 1470 and 1490, because the Δn(λ)/λ dependence of retardation serves to improve the stop-band width. Generally, a blue stage using retarder stacks with two or three individual retarder films gives acceptable coordinates.
Conversely, the Δn(λ)/λ dependence of retardation tends to reduce the stop-band width for the R/K stage 1440. Accordingly, additional retarder films are preferably used in the retarder stacks 1530 and 1550. For example, a 1.5λ red retarder film used in a retarder stack is 2.5λ in the blue, resulting in a magenta output. This can be addressed by using the LC modulator chrominance to eliminate the blue order, and by using the blue stage 1420 to further reduce blue transmission. However, this type of design is less robust because it relies on modulator retardance control and cooperation between stages to achieve acceptable spectra. As discussed above, it is preferable for each stage to be independent of the others. When 0.5λ red retarders are used, the blue order is safely eliminated, but the transition slope is shallower. While this allows the stages to be independent, the value of N (the number of individual retarders) must be increased to sharpen the transition slope and give an acceptable color coordinate.
While the DCS of
Asymmetric rise/fall times characterize the switching of nematic LC devices. A sufficient electric field applied across an LC film, which opposes the elastic forces, reconfigures the molecular director distribution in approximately 50-100 microseconds. When the field is removed, there is a relatively slow (approximately 500 microseconds to more than 1 millisecond) relaxation to the original distribution. Because this restoration is not electric field assisted, improvements in the reset time are accomplished by modifying the cell geometry, the surface treatment and the physical properties of the LC fluid. Surface mode devices are attractive because they provide sub-millisecond relaxation to the half-wave state, where other LC effects are either significantly slower or require very thin cell gaps. Surface mode devices can be driven to zero retardance in roughly 50 microseconds without difficulty.
The transient behavior of the preferred DCS design takes advantage of this difference in switching speed to maintain color coordinates, thereby eliminating the tradeoff between view time and saturation. For most ZTN modulator devices, maximum retardation is provided in the undriven state, with nearly zero in-plane retardation in the driven state. In the context of the preferred DCS design, this allows one to drive from any filtered state to a black state very rapidly.
As the stage corresponding to one primary color, e.g., green, is driven high, the modulator for the subsequent color, e.g., red is turned off. In less than 100 microseconds, the CSLM switches from green to black. Over the course of 1 millisecond, the retardance of the red modulator increases, allowing red transmission to increase to 100%. During this time, only the wavelength band established by the retarder stacks of the red stage is allowed to be transmitted, independent of the modulator retardance. As such, the color coordinate remains substantially constant as the transmission increases. It is only the relatively inconsequential chromatic behavior of the modulator over the additive primary band that can influence the color coordinate (at normal incidence). Using this scheme in a full-color sequential system, the CSLM (sequencers) can be viewed continuously, without blanking the display, with no impact on color quality.
There are also display modes in which the one-millisecond transition to the 100% transmission state can be beneficially used. In pulse width modulated displays, the resolution can be limited by the luminance step of a single frame. This step is determined by the minimum view time of a single frame under full illumination. A method for increasing resolution is to insert one or more frames of imagery with reduced illumination level, thereby increasing the view time to a manageable duration.
If there is a throughput benefit in decreasing the transition time between colors, alternative drive schemes may be considered. In a two-polarizer DCS, the interval between the driven state (black mode) and the undriven state (filtered mode) can determine throughput if the display is relatively fast. It is important to reach a half-wave retardance in the center of the operating band as quickly as possible, and to maintain that retardance throughout the view time.
Typically, the retardance in the low-voltage state is maintained at a particular value with the use of a small (2-4 volt rms) holding voltage. However, the effect of a holding voltage is to reduce the relaxation time to the half-wave retardance state. For instance, a particular modulator was found to have a 1.2 ms relaxation time (1.2 ms to go from, 10% transmission to 90% transmission between crossed polarizers) when operated with a 3 V rms holding voltage. The holding voltage was selected to provide the desired half-wave retardance throughout the 5.6 ms view time.
By using a three-level drive scheme, the relaxation time can be reduced. Initially, the modulator is switched from the maximum voltage state to a zero voltage state, which maximizes the relaxation to the half-wave state. After the desired retardance is achieved, the modulator can be driven with a small holding voltage for the remainder of the view time in order to maintain that retardance. The result is a 0.6 ms switching speed, corresponding to a 6% increase in throughput over that obtained with a two-level drive scheme.
In the four-polarizer multi-stage CSLM, neutral polarizers are inserted between each stage. Because the polarization is analyzed after each stage, the overall transmission is the product of the power transmission function of each stage. For this reason, a white state must be common to every stage. Note that this is not the case if dye based color polarizers are used rather than neutral polarizers. Unlike the two-polarizer case, cell chrominance of a stage in the white state does not effect the polarization introduced into other stages. This makes the filter less susceptible to color coordinate degradation from SA states and off-normal light. On the other hand, the addition of two polarizers does reduce overall throughput. Below we consider two four-polarizer examples; one using SO additive primary stages, the other using SO subtractive primary stages (or SA subtractive primary stages). We do not consider SA additive primary stages, due to the poor color coordinates obtained.
Note that every primary color that is output is the product of the color generated by one stage, and the white state from the remaining stages. In order to provide high throughput in the presence of significant modulator chrominance, it is necessary to tune the white state to provide maximum transmission at the center wavelength of the selected color. This affects the design of the modulators, since the half-wave retardance must be sufficient to provide a white state at the longest wavelength primary color that it accommodates. For example, the blue and green stages must give a half-wave retardance in the red, while the red stage must provide a half-wave retardance in the green.
There are additional benefits of tuning the white state. One benefit is the additional blocking provided by the low contrast filtering operations. For instance, additional blue rejection occurs from blue and green stages with their white states tuned to a red peak transmission. Conversely, the white state can be used in an analog mode to control color balance. A particular output primary color is substantially modulated by each of the other stages. As such, tuning the “white” stages from SA to SO modulates the desired output primary color between full transmission and zero transmission.
Another benefit of using SO additive primaries is that it is possible to generate a high density black state. Because each filtered state effectively blocks two additive primaries, there is a great deal of redundancy when each stage is in the SO mode. In fact, each additive primary band is blocked by two stages. If the modulators are implemented with nematic LCs, which have asymmetric switching rates, the beneficial transient behavior discussed above in connection with the two-polarizer DCS is exhibited. Specifically, the DCS can be driven rapidly to black, and the independence of stages provides high quality color coordinates during relaxation. Unlike the two-polarizer case, all three stages simultaneously generate additive primaries during the black output. The transmitted color is therefore determined by which combination of stages are switched to the white state. It is only the distortion of the transmission band caused by the evolution of the white states that can effect the color coordinate.
If the objective is to generate subtractive primaries, as is the case for certain electronic camera applications, the opposite of the above design considerations is true, except when a high density black state is also required. With parallel polarizers, cyan, magenta and yellow are generated independently by each of the three stages, and white can, in principle, be substantially lossless. However, the difficulty in obtaining a high density black state subtractively persists. Specifically, generation of the black state requires the cooperative action of C, M, and Y stages to achieve acceptable density. Even when cooperation between stages is optimized, the density of the black state is relatively poor. As such, the additive approach to producing subtractive primaries is usually preferred when a high density black state is required.
In this case, we obtain a DCS with behavior much like the two-polarizer SO subtractive primary DCS. Because each modulator modulates only one additive primary, the four-polarizer DCS can operate in either a SO subtractive or SO white mode with little difference in performance. The four-polarizer DCS is also suitable for implementing subtractive three-panel display systems.
Because the chromatic effects of the modulator are not significant, a high transmission white state can be generated in either SO subtractive primary or SO white modes. This eliminates the need to tune a chromatic white state. Using this design scheme, high quality subtractive primaries can be generated as well. However, the additive primaries can suffer from non-ideal cooperation between stages. The symptoms are much like those found in the two-polarizer DCS, including a true cyan side lobe in the red output, a true yellow side-lobe in the blue output, and a black state with relatively poor density. However, the exact behavior is different because the output in the four-polarizer DCS is the product of the transmission function of each stage. This means, for example, that stages cannot interact to improve rejection in the interprimary bands.
Selection of the modulator and the CSLM configuration is typically influenced by the field of view (FOV) requirement for a particular application. The FOV requirement can have a significant influence on the CSLM design and the modulator selection. In complementary color polarizer based systems, such as polarizer retarder stack (PRS), approximately equal weight is assigned to the FOV of the two voltage states, because both voltage states have a role in generating filtered spectra. Conversely, a system based on white/primary CSLMs permits unequal weights to be assigned to the voltage states. Specifically, a CSLM stage can be biased to generate a fixed color coordinate at large incidence angles in the filtered state, at the expense of the viewing angle in the white state. This design works best when a polarizer separates each stage, because such stages interact on a power transmission basis.
A benefit of the CSLM's of the present invention is that the modulator retardance does not represent a color shift. This is because, the transition bandwidths are defined by the retarder stacks and are independent of the state of the modulator. The modulator functions as a valve that determines the extent to which the spectral window established by the passive retarder stacks is filled.
Consider a four-polarizer three-color CSLM designed to give maximum FOV, as is required for sequential electronic camera applications. For this type of application, the switching speed is not critical, but a fixed color coordinate is desirable. Consider a CSLM structure that utilizes antiparallel aligned ZTN modulators. An anti-parallel aligned nematic LC cell gives a large field of view in the undriven state, because it consists only of a zero-order uniaxial in-plane half-wave retardance. Accordingly, there is no first-order shift in retardance with off-normal angles. Conversely, the driven state has a z-dependent director distribution that is not self-compensating. As such, the LC cell provides better FOV performance in the undriven state when uncompensated.
As discussed above, an RGB four-polarizer CSLM gives the best color coordinates using the SO filtered mode. Using an antiparallel nematic LC cell for the modulator means that the filtered output is assigned to the driven state, which has an inferior FOV. This problem can be solved by passively inverting the states. This is preferably accomplished by combining a negative birefringence in-plane uniaxial retarder film with the antiparallel nematic LC cell. The negative birefringence in-plane uniaxial retarder film is suitably polystyrene (Y. Fujimura et al., SID 92 Digest, p. 397) or a styrene derivative (U.S. Pat. No. 5,430,565). It could also be a preimidized polyamide or a liquid crystal polymer (H. Mori et al., European SID 97, M88).
A negative birefringence uniaxial in-plane half-wave retarder film oriented parallel to the antiparallel nematic half-wave retarder optic axis gives zero net retardance, as long as the dispersions are matched. Furthermore, the compound structure maintains zero retardance to very large angles of incidence and for all azimuths. Thus, the undriven state is converted from a half-wave retarder, with some sensitivity to incidence angle, to a zero-retardance that is substantially less sensitive to incidence angle. This gives the desired preference to the quality of the undriven state for the filtering operation.
Using this configuration, the color coordinate established by a stage in the filtering state is only marginally affected by the quality of the white states generated by the other stages. It is only the chrominance of the white states within the primary band that can shift the color coordinate. The effect is mainly to reduce the transmission of the primary, as the retardance shifts away from the half-wave retardance at the center of the band. This shift is enhanced by the relatively poor FOV of the driven state, in addition to the zero-order in-plane half-wave retardance from the film. In most cases, the associated loss in efficiency is much less objectionable than a color shift.
The retarder stacks can in principle be constructed of any transparent material that imparts differing optical path lengths to orthogonal polarizations. The typical range for color generation is 0.5-2.5 waves of visible retardance. Other desirable characteristics include low-cost, large area, uniform, thin, and durable under high illumination levels and harsh environments. Potential linear retarder technologies include form-birefringence structures, crystalline retarders, stretched polymer retarder sheets, and liquid crystal polymer films.
Form birefringence devices are based on periodic structures with features smaller than a wavelength. A grating with sufficiently high spatial frequency, for instance, does not diffract light from the zero order, and has optically distinct path lengths parallel and perpendicular to the grooves.
Conventional crystalline retarder materials, such as quartz, mica and calcite, are well suited to applications requiring higher resolution than is feasible with polymer films. They are also useful for applications requiring low wavefront distortion, and/or high power handling requirements. They are more expensive than polymer retarders and do not lend themselves to large area, particularly when low retardances are required.
Polymers are chemical compounds or mixtures of compounds consisting of repeating structural units formed by a chemical reaction, where two or more small molecules combine to form larger molecules. Retarders can be formed using sheets of oriented organic polymers. By unidirectional or bidirectional stretching of polymer films at elevated temperatures, a linear birefringence can be induced. For most materials, such as polyvinyl alcohol or polycarbonate, the refractive index increases in the direction of the stretch, typically forming a positive in-plane uniaxial retarder. Materials such as polystyrene, and styrene derivatives (U.S. Pat. No. 5,430,565) give a negative in-plane birefringence after stretching. Such materials are useful for producing wide-view angle CSLM structures. Retardation films exhibiting fixed retardation with off-normal view angles are produced using one of two techniques, biaxial stretching or compound retarders. The combination of a positive uniaxial film crossed with a negative uniaxial film of equal retardation results in a compound retarder with an isotropic, angle-insensitive retardation (U.S. Pat. No. 5,138,474).
As an alternative, a single layer biaxial film with wide view angle can be fabricated using stretching along both x and z, as described in U.S. Pat. Nos. 5,245,456, U.S. Pat. No. 5,472,538, and U.S. Pat. No. 5,559,618. Such stretching can produce a biaxial film with a z direction refractive index that is substantially intermediate between the in-plane refractive indexes. This single film behaves much like the two-layer film in providing a large isotropic view angle.
Stretched polymer films are available in arbitrary retardances topically from 0-2,000 nm, using a variety of materials with unique birefringence dispersion characteristics. Large sheets can be purchased at low cost, permitting large clear aperture filters. Polycarbonate is the most common material for display applications and is available from vendors such as Nitto Denko, Sumitomo, and Sanritz. Other materials, such as poly-vinyl alcohol, can be purchased from Polaroid, Polarvision, Inc., and International Polarizer. High dispersion materials, such as polysulphone and polyarylate films, are available from Nitto Denko and Sumitomo. Several other polymer materials are potentially useful in producing stacks including, but not limited to, mylar, polyamide, polypropylene, polyethylene terapthalate, triacetate (tri-butyl-acetate), polytetrafluoroethylene, and polymethylmethacrylate.
As discussed above, spatially patterned CSLM structures, as for color filter arrays, can be implemented using spatially patterned polymer retarders. Here, a film with a spatially patterned retardation replaces or works in concert with the active LC element. Such patterned retarder films can be fabricated by locally stressing an isotropic polymer film. Half-wave retarder stripes can be patterned on an isotropic substrate by locally stressing a film.
Liquid Crystal Polymers (LCP) are a class of polymers wherein liquid crystal monomers are incorporated into the macromolecular structure along the mainchain (backbone) or as side chain units. LCP's can be aligned by either mechanically rubbed surfaces, shearing, or by optical means. Optical methods involve first applying either linear photo-polymerizable (LPP) films or azo-based dyes either in a polymer alignment layer.
If LPP films are used, (see Schadt, et al, Jpn. J. Appl. Phys. Vol. 34, pg. 3240-3249, 1995), the LPP materials are deposited on a substrate and then cured at elevated temperature. The cured film is then subjected to polarized UV light. LCP's are then spun-on or coated onto the same substrate, and aligned with the orientation of the LPP film. The LCP's are then cross-linked by exposure to unpolarized UV light.
If azo-based dyes are used, (Shannon et al, Nature, vol. 368, pg. 532-533, 1994), azo-dye molecules are intercalated into the polymide alignment layer (or layers), which are deposited onto various substrates, e.g., glass, silicon, or others. A liquid crystalline monomer or polymer is either deposited onto one substrate, or sandwiched in between two substrates. The LC molecular director orients perpendicular to the direction of the polarized UV light which previously illuminated the alignment layer. Subsequent exposure will reorient the liquid crystals, which may be disadvantageous for some applications.
Liquid crystal polymer films, particularly UV cross-linkable polymer nematic retarders, are particularly suitable for forming retarder stacks. An attractive feature is the ability to produce thin retarders, since the material can have high birefringence relative to stretched materials. This can permit the fabrication of multi-layer retarder stacks on a single substrate, such as the polarizer substrate, with low cost. Because the films can be patterned at arbitrary angles, there is no waste, as is the case when cutting stretched polymer films at angles. Each LCP layer can essentially be bonded to the previous layer, avoiding the need for applying pressure sensitive adhesives to each film. In addition, LCPs are conducive to forming complex impulse response retarder stacks, because the retardation of each LCP layer is arbitrary. Such a capability increases the degrees of freedom, which results in greater design options. Conversely, there are significant cost pressures to minimize the number of unique retardation values in each retarder stack when using stretched polymer films.
The retarder stacks discussed above preferably comprise discrete layers, i.e., individual retarders, with specific orientations and retardance values. However, continuous retarder “stacks” or structures can be fabricated using layers of LCP with twisted structures. Thus, the retarder stack is not limited to a discrete number of individual zero-twist (linear) retarders. By selecting film thickness, birefringence, and the orientation of molecules at the boundary of each surface, films self-assemble into continuous retarder structures. Thin glassy films can be formed on a substrate using LCP materials. Such films are equivalent to a large number of very thin individual linear retarders, but avoids the need to physically stack individual retarder layers. Other color options are available by stacking twisted LCP layers or twisted and zero-twist LCP layers.
Currently available LCP materials have a positive anisotropy. In order to produce a wide view angle LCP, a biaxial retarder is required. A single film would require a true biaxial, LCP (P. Bladon et al., Macromolecules, 25, 4329 (1992)). An alternative is a two-layer compound retarder containing positive and, negative birefringence LCPs. Yet another alternative is to take advantage of the out-of-plane tilt achievable with photoaligned LCP. By providing two identical parallel oriented LCP films with equal and opposite out-of-plane tilt angle, a compound retarder is formed with a view angle that is controlled by the tilt angle. At roughly 34-degrees, the off-normal dependence of retardation can be made to vanish in the plane containing the optic axis.
Because LCPs can be photoaligned, a spatially varying orientation, both in-plane and normal to the surface, can be achieved by photo-patterning. This is done using standard lithographic techniques. As such, spatially patterned versions of CSLM structures can be produced that behave as color filter arrays. Here, the active modulator of a CSLM is replaced by a spatially patterned LCP retarder that locally determines the transmitted color. In addition, hybrid structures using both spatially-switched and actively-switched layers can be incorporated. In one embodiment, patterned retarders are used in series with the active modulator, so that modulation between fixed spatial patterns can be accomplished using an LC modulator device with unpatterned electrodes.
Other LCP materials are potentially useful for either retarder stack materials or for increasing the view angle. Discotic LCPs from Fuji and Nippon Oil Company are negative birefringence materials fabricated on a CTA substrate (H. Mori, Jpn. J. Appl. Phys., 36, 1068, (1997)). Some of the Discotic LCPs have the unique property that the optic axis has a splayed orientation that approximates that of the active modulator. This is quite useful for compensating the select state of, for instance, a TN device.
Other useful materials for increasing view angle are negative birefringent films with optic axes oriented along z. For instance, the cellulose triacitate (CTA) substrate of typical iodine display polarizers has 40-70 nm of retardance normal to the substrate on either side of the polarizer. Larger retardances can be obtained by biaxial stretching of polymers, such as polycarbonate (Nitto Denko and Sumitomo). Typically, such films exhibit a small positive in-plane retardation. Other materials, such as preimidized polyamide, are negative birefringent materials that align with their optic axis normal to the substrate, giving zero in-plane retardance.
Multiple films of negative birefringent film can be combined to form a composite structure that can fully compensate one state of the LC modulator. Preferred LC modulators, such as the pi-cell, are those in which the passive structure improves the view angle of both the select and non-select states. Often, however, compensation improves field of view of one state at the expense of the other. This represents an advantage of CSLM structures, because the state providing the largest view angle can be assigned to the filter state. The compromised state is thus the neutral state, which can be made to have little influence on the color coordinate.
In the embodiments described and shown, the CSLM is illustrated with linear polarizers. In general, they can be linear, circular or elliptical polarizers. They can be neutral, or can exploit a wavelength-dependent polarization. Polarizer materials are available in a number of forms, the selection of which depends upon the application requirements. Suitable polarizers include absorption-based, e.g., crystal dichroism long chain molecules of dye or iodine and oriented metal particles (Polarcor product, Corning, Inc.), polymer multilayers, e.g., DBEF product, 3M, Inc., birefringent, e.g., calcite or quartz-based Ahrens, Wollaston, Rochon, Nicol, or Glan polarizers, scatter polarizers using rough surfaces (Yamaguti) or LC in a polymer matrix, reflection, e.g., cholesteric LC, multilayers on 45-degree prisms, coated prismatic, and periodic structures, e.g., coated gratings, slits, wires, or oriented silver or copper particles (Polarcor product, Corning Inc.).
Dichroic sheet polarizers are thin, low cost, large area, are easily integrated, and can perform well at large angles. High efficiency iodine polarizers, such as Nitto Denko EG1225 are appropriate for low power applications. Durable high luminance polarizer films for projection are typically based on dye-stuff, e.g., Polatechno and Polaroid KE product, with reduced throughput (84%). For normal incidence, the polarizer density need not be dramatically higher than is theoretically feasible based on side-lobe leakages from the retarder stack. As such, density can be traded for increased throughput if the polarizer technology permits.
In wide view-angle applications, polarizer FOV must be considered. Light leakages in the ±π/4 azimuths, relative to the polarizer axis, are most pronounced. Due to geometrical effects, crossed polarizers cease to appear crossed in these azimuths at off-normal incidence. In addition, substrate retardance can contribute to a loss in density. Based on these considerations, polarizers such as HN22 (Polaroid), which exhibit high density and use a cellulose acetate butyrate (CAB) substrate, are preferred for large incidence angles. Methods for compensating geometrical effects and substrate retardance effects can also be implemented (J. Chen et al., SID '98 Digest, 315-318).
CSLM structures can be fabricated using techniques that are standard to the industry. Liquid crystal cells (modulators) are preferably fabricated using float glass or borosilicate glass panels, though plastic cells are also suitable. Borosilicate, such as Corning 7059, provides higher transmission than certain float glasses. The glass is coated with a transparent conductor for addressing the liquid crystal. A good conductor is indium tin oxide (ITO), typically with a resistivity of 100-1,000 Ω/cm2. By coating oxide films between the glass and the ITO, such as Ta2O5/MgF2/ITO, reflection losses can be minimized. In so doing, modulator transmission can be increased by several percent, often yielding upwards of 99% transmission.
An alignment layer is coated on the transparent conductor for orienting the LC film. Standard materials are polymides, such as those from Nissan Chemical, though photoalignment materials (LPP) can also be used. The former is mechanically rubbed after processing, while the latter is exposed to UV light.
An adhesive gasket is dispensed or screen printed, typically with a 0.5-1.5 mm width around the perimeter of the cell, leaving a small fill hole. The glass substrates are mated and the adhesive is cured under pressure using either UV light, temperature, or a sequence of both. Spacers of glass rods or spheres, or flexible polymer balls are used to determine the spacing between conductors. Requirements for spacer density depend upon uniformity requirements, glass thickness/flatness, aperture size, and liquid crystal EO effect. Subsequently, the panel is broken into sticks for filling. The liquid crystal material is back-filled in the space between the conductors, and aligns according to the surface treatment used. In some cases, a filled part undergoes a post processing step to improve uniformity. The fill hole is then plugged using an adhesive. Contact is made to the ITO along one or two ledges (depending upon whether cross-over dots are used). Conductors typically include soldered wires, or pins.
Preferred retarder stacks are constructed from biaxial polycarbonate films and, when needed, neutral polarizer films. Typically, vendors supply both polarizer and retarder materials, and are therefore capable of building the complete retarder stack. Using pressure sensitive adhesives (PSAs), the complete retarder stack can be laminated together into a single unit, which is supplied with a protective layer on the polarizer and a release sheet on the retarder PSA. The retarder stacks can also be supplied die-cut to the appropriate product size.
Retarder stacks can be applied to the external surfaces of the LC modulator using the PSA, forming a single CSLM stage. Multiple stage CSLMs can be fabricated by cementing the stages together using standard materials. These includes UV, thermal, pressure cure epoxies, RTVs, and an adhesive gasket material with a back-filled index matching material. Should plastic LC devices be used for the modulator, the entire CSLM can be fabricated using lamination techniques with PSAs. Wavefront distortion of the assembled unit can be compromised by the texture of the external polymer surfaces. This problem is greatly alleviated by cementing a glass end-cap over the external polymer surface. Furthermore, transmission can be improved by using an end-cap which is antireflection coated.
Transmission losses in retarder stacks are the result of several loss mechanisms. Typically, the retarder stacks are comprised of several sheets of polycarbonate with acrylic based PSA's between them. Losses are thus a consequence of material absorption and random scatter, as well as reflections at the interface between each PSA and polycarbonate layer. Reflection at the interface between each PSA and polycarbonate layer is typically dominant, as PSA material often has a refractive index far below that of the polycarbonate. Thus, methods for increasing transmission focus on eliminating surface reflections. One method of eliminating surface reflections is to use a PSA that is better matched to the polycarbonate refractive index. Another method is to use chemical bonding. In chemical bonding, the polycarbonate sheets are welded together using a suitable solvent. For example, polycarbonate dissolved in methylene chloride provides an index matched material.
As discussed above, there are generally four ways to implement color displays: (1) spatial; (2) sequential; (3) split-path; and (4) common-path. The CSLM can be used purely as a color sequencer in a sequential architecture, or can be used to implement a common-path display in a stacked-panel architecture. In addition, hybrid architectures can be implemented that use CSLM devices in displays that combine any of the above four color generation techniques.
A full color sequential display (design #2 in
The light source 1500 is suitably a metal halide lamp and preferably emits optical power in all three primary color bands. Alternatively, the light source 1500 can be implemented with an active lamp system or with a lamp/color wheel combination.
For illustrative purposes, the two-polarizer digital polarizer color sequencer 1550 is of the type shown in
The blue stage 1420 comprises a first retarder stack 1470, a blue LC modulator 1480 and second retarder stack 1490. The green stage 1430 comprises a first retarder stack 1500, an LC modulator 1510 and a second retarder stack 1520. The red stage 1440 comprises a first retarder stack 1530, an LC modulator 1540 and a second retarder stack 1550. The input and output polarizer 1450 and 1460 are crossed.
In operation, the light source 1500 and the sequencer 1550 sequentially illuminates the liquid crystal display 1560 with red green and blue light. The liquid crystal display 1560 is sequentially driven with red green and blue image information in synchronism with the red green and blue illumination from the light source 1500 and the color sequencer 1550. The liquid crystal display 1560, in combination with the polarizer 1570, modulates the intensity of the light that is sent to the screen 1590 in accordance with the image information. The full color sequential display of
As discussed above, other types of illumination systems can be used in place of the light source 1500 in the displays of
In the display shown in
By placing the digital color sequencer between the light source 1500 and the liquid crystal display 1560 or 1600, the image at the screen 1590 is not sensitive to any wave-front distortion caused by the digital color sequencer 1550.
The digital color sequencers 1550 can be made relatively small, even though the power handling is relatively high when the digital color sequencer 1550 is positioned as shown in
Alternative locations for the digital color sequencer 1550 are between the liquid crystal display 1560 or 1600 and the projection lens 1580, and between the projection lens 1580 and the viewing screen 1590. If the digital color sequencer 1550 is located between the projection lens 1580 and the viewing screen 1590, the digital color sequencer 1550 must have a relatively large clear aperture and low wave-front distortion. In addition, this location will reduce the power handling requirements of the digital color sequencer.
For a high-frame rate system, the LC display 1560 or 1600 is preferably a silicon-based display, rather than a TFT. Typical reflective LCDs integrate ferroelectric LC, or a variety of nematic LC modes, between a glass substrate coated with a solid ITO electrode and a reflective active matrix structure on a silicon substrate. In such systems, the analyzing polarizer of the DCS 1550 can serve as the input polarizer to the LCD panel, or vice versa. In a system that uses off-normal illumination, such as a DMD the sequencer can be placed in either the illumination or projection arms of the system. However, in a retroreflecting arrangement, a beam splitting polarizer functions both to create a crossed polarizer and to route the beam through the system. In such a configuration, the sequencer can also be inserted in the arm between the beam splitting polarizer and the LCD. However, additional constraints are imposed on the design of such a CSLM based color sequencer.
The foregoing embodiments are merely exemplary and are not to be construed as limiting the present invention. The present teaching can be readily applied to other types of apparatuses. The description of the present invention is intended to be illustrative, and not to limit the scope of the claims. Many alternatives, modifications, and variations will be apparent to those skilled in the art. In the claims, means-plus-function clauses are intended to cover the structures described herein as performing the recited function and not only structural equivalents but also equivalent structures.